Australasian guidelines for antifungal therapy in - Mycology Online

27 Pages·2008·0.25 MB·English

Australasian guidelines for antifungal therapy in neonates and children
with proven, probable and suspected invasive fungal infections

CC Blyth, TA O’Brien and P Palasanthiran on behalf of the Australasian Society of
Infectious Diseases (Paediatric Infectious Diseases and Mycoses Interest Groups) and
Bone Marrow Transplantation Committee (Australian and New Zealand Children’s
Haematology/Oncology Group)

Dr Christopher C Blyth
MBBS
Department of Immunology and Infectious Diseases, Sydney Children’s Hospital,
High Street, Randwick. NSW, Australia
School of Women’s and Children’s Health, University of New South Wales,
High Street, Randwick, NSW. Australia

Dr Tracey A O’Brien
FRACP, MBChB, MHL, BSc
Centre for Children’s Cancer and Blood Disorders, Sydney Children’s Hospital,
High Street, Randwick. NSW. Australia
School of Women’s and Children’s Health, University of New South Wales,
High Street, Randwick, NSW. Australia

Dr Pamela Palasanthiran
MBBS FRACP MD
Department of Immunology and Infectious Diseases, Sydney Children’s Hospital,
High Street, Randwick. NSW. Australia
School of Women’s and Children’s Health, University of New South Wales,
High Street, Randwick, NSW. Australia

Corresponding Author:
Dr Pamela Palasanthiran
Email: [email protected]

Conflicts of interest: nil

(Version: 2007)
Finalised version: September 2007 1
Abstract
The management of children with invasive fungal infection is becoming more complex with an
increasing number of antifungal agents available. Insufficient paediatric studies have been
performed with newer agents leaving paediatricians reliant on adult patient data to guide clinical
decisions. There are numerous differences between children and adults with invasive fungal
infection. Following a systematic review, consensus guidelines have been developed to assist
doctors managing children with invasive fungal infection. The efficacy, toxicity and cost of newer
antifungal agents have been compared with existing therapies and key recommendations made.

Introduction
The incidence of invasive fungal infection is increasing, especially in at risk populations.1;2 Risk
factors for invasive fungal infection in children include malignancy, haematopoietic stem cell or
solid organ transplantation, neutropenia, chemotherapy, corticosteroid use, congenital and
acquired immunodeficiency, prematurity, broad spectrum antimicrobial use and dependence on
parenteral nutrition.3-5 Invasive fungal infections are responsible for significant morbidity and
mortality. In studies published since 1996, mortality in children with candidaemia ranged from 19
to 32%.6-10 Invasive aspergillosis in children is associated with even greater mortality: 68 to
77%.3;5;11;12 Mortality is highest in those with greater degrees of immunosuppression, particularly
following haematopoietic stem cell transplantation.12

Australasian treatment guidelines for adults with invasive fungal infection have been formulated
with updated guidelines currently being developed.13 Significant differences between children and
adults with invasive fungal infection are reported including predisposing factors, infecting
organism and site of infection.3;5;14 There are also significant differences between paediatric and
adult antifungal pharmacokinetics and toxicity.15-18

The number of drugs available to treat fungal infections has greatly increased in the last decade.
Four classes of drugs for the treatment of invasive fungal infections exist: polyenes, triazoles,
echinocandins and nucleoside analogues. In Australia and New Zealand, conventional
amphotericin B (CAB), liposomal amphotericin B, amphotericin B lipid complex (ABLC),
amphotericin B colloidal dispersion (ABCD), fluconazole, itraconazole, voriconazole,
posaconazole, caspofungin and flucytosine are available for use in invasive fungal infection. New
azoles and echinocandins are expected to be available in the near future.

Finalised version: September 2007 2
This set of paediatric guidelines is based on the best available current evidence (Box 1).
Standard definitions were used (Box 2). The guidelines aim to promote rational prescribing of
antifungal agents for Australian and New Zealand children with proven, probable and suspected
invasive fungal infections. When paediatric studies were judged by the authors to be inadequate,
adult studies have been used to supplement data. Antifungal susceptibility profiles (Box 3),
paediatric antifungal dose guidelines (Box 4) and key recommendations (Box 5) are given.

Spectrum of antifungal activity
Antifungal susceptibility patterns of the organism confirmed or suspected should be used to guide
antifungal therapy. Antifungal susceptibilities of common fungal pathogens are summarized in
Box 3.19 Amphotericin B preparations are active against most fungal species. C. lusitaniae is
reported to have variable susceptibility to amphotericin B, however this has not been found in
Australia. Scedosporium and Fusarium species have reduced susceptibility to amphotericin B and
it is common practice to use high doses of amphotericin B (ie ≥ 5mg/kg of a lipid preparation) for
infections due to zygomycetes.19

Fluconazole has activity against many yeasts but limited activity against moulds.19 Reduced
susceptibility to fluconazole is also seen with C. krusei and C. glabrata. C. tropicalis has been
reported to have reduced susceptibility to fluconazole in international studies but this is rarely
seen in Australia. Itraconazole has anti-Aspergillus activity and variable activity against other
moulds. Voriconazole has activity against most yeasts and moulds; zygomycetes and
Scedosporium prolificans are important exceptions. Zygomycetes are susceptible to
posaconazole. S. prolificans is resistant in vitro to all available antifungal drugs.

Candida and Aspergillus species are susceptible to caspofungin.19 Caspofungin has poor activity
against Cryptococcus neoformans, Scedosporium species, Fusarium species and zygomycetes.

Empiric antifungal therapy in prolonged fever and neutropenia
Summary
• No differences in treatment success and mortality have been demonstrated in paediatric and
adult trials comparing empiric conventional amphotericin B and lipid preparations in subjects
with prolonged fever and neutropenia [I].
• Fewer breakthrough infections may occur in subjects treated with liposomal amphotericin B [I].
• Conventional amphotericin B and fluconazole appear equally effective in febrile neutropenic
subjects not previously administered fluconazole [II].
• Voriconazole is not inferior to liposomal amphotericin in the management of adults and
adolescents with fever and neutropenia although equivalence has not been demonstrated [II].
Finalised version: September 2007 3
• Caspofungin appears as effective as liposomal amphotericin B in the treatment of febrile
neutropenic adults [II].

Despite the limitations of the original studies,20;21 it is the accepted standard of care to use
antifungal agents in neutropenic subjects who remain febrile despite broad spectrum antibacterial
agents.22

Paediatric trials comparing antifungal agents in prolonged fever and neutropenia are limited.
Prentice et al compared liposomal amphotericin B (1 or 3 mg/kg/day) with CAB (1 mg/kg/day) in
202 neutropenic children with 96 hours of fever despite appropriate therapy.23 Safety was the
primary endpoint. Treatment success was not significantly different between groups: 51% in
children receiving CAB, 64% in those receiving 1mg/kg/day liposomal amphotericin B and 63% in
those receiving 3mg/kg/day liposomal amphotericin B (p = 0.22) [II].

Therapeutic trials in adults also report no difference in treatment success when CAB (0.6 - 1
mg/kg/day) is compared with liposomal amphotericin B (3 mg/kg/day) or ABCD
(4mg/kg/day)[II].23-25 Two meta-analyses failed to demonstrate any difference in mortality between
CAB and lipid preparations of amphotericin B in patients with fever and neutropenia (Johansen et
al. OR: 0.83, 95%CI: 0.62 – 1.12; Barrett et al. OR: 0.78, 95%CI: 0.54 – 1.13) [I].26;27 Johansen et
al demonstrated fewer breakthrough fungal infections in those receiving liposomal amphotericin B
compared with CAB [I].26 Barrett et al also included antifungal therapy in neutropenic and non-
neutropenic populations with confirmed candidiasis, aspergillosis and cryptococcosis. In the
combined analysis, all cause mortality was reduced in those receiving a lipid preparation (OR:
0.72, 95%CI: 0.54 - 0.97) but no difference in efficacy was demonstrated (OR: 1.21, 95%CI:
0.98 – 1.49).

Of the trials comparing fluconazole and itraconazole with CAB in neutropenic patients, only one
was appropriately powered and only two studies included children.28-32 All subjects were azole
naïve. Azoles were as effective as CAB [II]. Voriconazole was compared with liposomal
amphotericin B in paediatric and adult subjects. The overall success rates were 26.0% with
voriconazole and 30.6% with liposomal amphotericin B (95% confidence interval for the
difference, -10.6 to 1.6%) [II].33 There were fewer documented breakthrough fungal infections in
patients treated with voriconazole than in those treated with liposomal amphotericin B. Walsh et
al compared liposomal amphotericin B with caspofungin in febrile neutropenic adults. Despite a
greater number of caspofungin-treated patients surviving for ≥ 7 days after study completion, no
overall difference in treatment success was observed [II].34

Finalised version: September 2007 4
Antifungal therapy in candidaemia or invasive candidiasis
Summary
• No difference in treatment success is seen when conventional amphotericin B is compared
with fluconazole and voriconazole in non-neutropenic adolescents and adults [II].
• Echinocandins are as effective as conventional amphotericin B and fluconazole in non-
neutropenic adults [II].
• Prior antifungal therapy influences the infecting Candida species and should guide empiric
therapy [III-2].
• Therapy in candidaemia should be continued for at least 14 days following the last positive
culture [expert opinion].
• If feasible, initial management should include removal of all existing vascular catheters [III-2].

Candida albicans and C. parapsilosis are the most frequent yeasts isolated in children with
invasive candidiasis. Species with reduced susceptibility to fluconazole are infrequent pathogens
in children.1;19;35-37 Prior fluconazole therapy is a risk factor for infection with a fluconazole
resistant yeast [III-2].37;38 A strong association between C. parapsilosis and central venous
catheter infection has been documented [III-2].35;39

Randomised controlled trials comparing fluconazole with CAB or itraconazole in paediatric
candidaemia are insufficiently powered to assess treatment efficacy.40;41 The therapeutic
equivalence of fluconazole and CAB has been demonstrated in predominantly non-neutropaenic
adolescents and adults with candidaemia [II].42-44 No significant differences in efficacy between
fluconazole and voriconazole are seen in adolescents and adults with candidaemia and
oesophageal candidiasis [II].45;46 Caspofungin and CAB appear equally effective in adult patients
with candidaemia and oropharyngeal/oesophageal candidiasis [II].47-49 Therapeutic equivalence is
also seen when fluconazole is compared with caspofungin in adults with oesophageal candidiasis
[II].50 Little data comparing antifungal therapy in deeper candidal infection exists. An amphotericin
preparation (with flucytosine where possible) is favoured as initial therapy in candidal endocarditis
and meningitis [IV].51

Therapy for 14 to 21 days after the last positive blood culture is recommended in children and
neonates with candidaemia in the absence of disseminated disease [expert opinion].51 Some
authors suggest that 7 to 14 days of therapy following blood stream sterilisation can be used in
children. This recommendation is supported by data from a single case series [IV].52

Finalised version: September 2007 5
Intravascular devices are a frequent source of candidaemia. In patients with severe neutropenia
and/or receiving chemotherapy, candidaemia is frequently of gastrointestinal origin.53 Studies in
non-neutropenic adults, children and neonates demonstrate that early catheter removal is
associated with a shorter duration of candidaemia and reduced mortality [III-2].54-56 Early removal
is often not possible in unstable patients or in the presence of implantable vascular access
devices. If feasible, initial non-medical management should include removal of all existing
vascular catheters [expert opinion].51

Antifungal therapy in invasive aspergillosis
Summary
• There are no adequately powered randomised controlled trials comparing conventional
amphotericin B with lipid preparations in probable or proven invasive aspergillosis.
• In pulmonary aspergillosis, 3mg/kg/day of liposomal amphotericin B is as effective as
10mg/kg/day [II].
• Initial therapy with intravenous then oral voriconazole is superior to conventional
amphotericin B in invasive aspergillosis and is associated with a survival advantage [II].
• Caspofungin is an alternative in subjects intolerant of, or refractory to, other therapies
[IV].
• The duration of therapy should be guided by the clinical and mycological response and
immunological recovery [expert opinion].

No randomised controlled trials have been performed in children with invasive aspergillosis.
Using comparative data from adults, a number of conclusions can be drawn regarding the relative
efficacy of different antifungal agents. CAB was compared with liposomal amphotericin B and
ABCD in two small randomised controlled trials. No significant difference in treatment outcome
was observed [II].57;58 High dose liposomal amphotericin B (10mg/kg/day) was compared with
standard dose (3mg/kg/day) in 339 adults with filamentous fungal infections (predominantly
pulmonary aspergillosis). No differences in treatment success at 12 weeks (50% vs. 48%) or
survival (72% vs. 58%) were observed [II].59

Herbrecht et al compared CAB (1 - 1.5mg/kg/day) with voriconazole (6mg/kg bd for 24 hours then
4mg/kg bd) in 277 adolescents and adults with proven or probable aspergillosis.60 Therapy was
initiated with intravenous voriconazole for at least 7 days. Treatment success (53% vs 32%) and
survival (71% vs 58%) was greater with voriconazole. The mean duration of study drug was
longer in those administered voriconazole (77 days vs. 10 days). Alternative licensed antifungal
agents were allowed and were more frequently prescribed in those receiving CAB (80% vs. 36%).
The impact of this on treatment success and survival is unknown.
Finalised version: September 2007 6
Echinocandins were effective in adults with aspergillosis who were intolerant of or refractory to
other antifungals [IV].61;62 No comparative trials have been performed.

The duration of antifungal therapy for invasive aspergillosis in children and adults has not been
determined. Herbrecht et al evaluated subjects clinically and radiologically following 12 weeks of
antifungal therapy.60 In an open label study of voriconazole use in children with invasive fungal
infection (predominately aspergillosis), the median duration of therapy was 93 days (range 1 –
880).63 The length of therapy should be influenced by response to therapy and immunological
recovery.

Combination antifungal therapy in children with candidaemia, invasive
candidiasis or aspergillosis
Summary
• Insufficient evidence is available to support routine use of combination therapy in
candidaemia, invasive candidiasis or aspergillosis.
• Two antifungals from different classes are recommended if combination therapy is used.

Given the high mortality of IFI, the role of combination therapy has been considered. To date, no
paediatric combination trials have been published although research is underway. In 219 non-
neutropenic adults, persistent candidaemia after 5 days of therapy was more frequently seen with
fluconazole alone compared with fluconazole and CAB combination therapy.64 No statistically
significant difference between the groups was seen in treatment success at 30 days and 90 day
mortality. Interpretation is further confounded by differences in severity of illness between study
groups. The combination of voriconazole and caspofungin were retrospectively compared with
voriconazole alone as salvage therapy in a unmatched population with proven or probable
aspergillosis.65 Three month survival was superior in subjects receiving combination therapy.

Other fungal pathogens
There are insufficient data on children to assist paediatricians managing infections caused by
fungi other than Candida and Aspergillus spp. Therapy should be guided by known antifungal
susceptibility patterns (Box 3), underlying disease, site and extent of infection.

The following antifungal agents are recommended for treatment of cryptococcosis,
scedosporiosis, fusariosis and zygomycosis:

Finalised version: September 2007 7
• Cryptococcus neoformans complex: An amphotericin B formulation is recommended in
the induction phase of therapy for cryptococcal infection. The addition of flucytosine is
recommended in those with central nervous system infection, HIV infection and children
receiving immunosuppressive therapies [II].66-69 Induction therapy should be followed by
maintenance therapy, usually with fluconazole. Fluconazole is an alternative agent in
patients with localised disease due to C. neoformans var. neoformans (eg pulmonary and
skin disease). Fluconazole is also the agent of choice for long term suppression in the
setting of continual immunosuppression [II].
• Pseudallescheria boydii (Scedosporium apiospermum) and Scedosporium prolificans:
Voriconazole is the preferred agent for P. boydii and S. prolificans infection [IV].13;63;69;70
The addition of terbinafine for synergy is recommended in subjects with S. prolificans
infection [IV].13;69;71
• Fusarium species: An amphotericin B preparation72 or voriconazole73 is the treatment of
choice for fusariosis [IV].13;69;74
• Zygomycete species: High dose lipid preparation of amphotericin B75 in conjunction with
surgical debridement is currently recommended for zygomycetes [IV].13;69;74 The role of
posaconazole is still to be determined.76

Salvage fungal therapy
There is no consensus on the definition of refractory fungal infection. In candidemia, treatment
failure is defined by Rex et al as persistent fungemia or unimproved / progressive signs of sepsis
despite ≥ 5 days antifungal treatment.43;64 Almyroudis et al propose that a composite outcome be
used to define refractory invasive mould infections. Refractory infection is defined as worsening of
at least 2 of 3 criteria: clinical, radiologic and mycologic, with assessment after a minimum of 7
days of therapy.77 It should be noted that the radiological appearance of pulmonary aspergillosis
often worsens despite adequate antifungal therapy and that immune reconstitution may be
associated with worsening of radiological signs.78

Alternative antifungal therapies for candidaemia, invasive candidiasis and aspergillosis are
included in the key recommendations (Box 5). If invasive mould infection is suspected, and the
infecting species is unknown, voriconazole would be the preferred choice of antifungal given its
broader spectrum of activity. The role of combination therapy is yet to be established.

Toxicity of antifungals
Summary
• Lipid preparations are less nephrotoxic than conventional amphotericin B when
administered as a standard infusion in studies comprising predominantly adults [I].
Finalised version: September 2007 8
• Conventional amphotericin B administered by continuous infusion results in less
nephrotoxicity and infusion-related toxicity compared with a standard infusion [II]. Other
agents have not been compared with continuous infusion.
• Increasing cumulative doses [III-2], preexisting renal impairment [III-2], hypovolaemia [II],
hyponatraemia [II] and the number of concurrent nephrotoxins [II] are associated with an
increased risk of amphotericin B induced nephrotoxicity. Concomitant use of cyclosporine
or diuretics with amphotericin B increases the risk of nephrotoxicity [III-2].
• Haematopoietic stem cell transplant recipients are the highest risk group for
nephrotoxicity and dialysis following amphotericin B administration [III-3].
• Azoles and echinocandins are less nephrotoxic than conventional and liposomal
amphotericin B [II].
• Liposomal amphotericin B causes less infusion-related toxicity when compared with
conventional amphotericin B and ABLC [II].
• Both azoles and echinocandins cause less infusional toxicity than conventional and lipid
preparations of amphotericin B [II].
• The incidence of hepatotoxicity is similar across all antifungal classes [II].
• Significant drug interactions can occur with azoles requiring dose modification or
avoidance [IV].

The development of new antifungals has been driven by the toxicities associated with CAB.
Amphotericin B preparations can cause renal insufficiency by a number of mechanisms. This is
usually reversible, although renal dysfunction may remain for weeks upon cessation.45;79
Permanent damage has been reported.80;81 In adults, renal toxicity is associated with increased
mortality, increased length of stay and total costs.82 These outcomes have not been
demonstrated in children.

Published rates of CAB induced nephrotoxicity in children vary widely; 1.2 - 52%.23;25;83-85 The
different definitions of nephrotoxicity make comparison difficult. Nephrotoxicity has been
examined in three randomised controlled trials comparing CAB with lipid preparations in children.
Comparing CAB (0.8mg/kg/day) with ABCD (4mg/kg/day), Sandler et al and White et al defined
nephrotoxicity as a doubling of baseline creatinine, an increase in serum creatinine of 88umol/L
or a 50% decrease in creatinine clearance. Both trials demonstrated reduced nephrotoxicity with
ABCD: 52% vs. 12%, p < 0.01 [II].25;84 Comparing CAB (1mg/kg/day) to liposomal amphotericin B
(1-3mg/kg/day), Prentice et al defined nephrotoxicity as a doubling of baseline creatinine. The
rate of nephrotoxicity was 21% vs 8-11% but failed to achieve statistical significance with p =
0.10.23 None of the trials identified the proportion of children in whom serum creatinine exceeded
the upper limit of normal or another predefined measure of renal impairment.
Finalised version: September 2007 9
Two meta-analyses of trials including mostly adult patients demonstrated a 49-75% reduction in
nephrotoxicity in subjects administered lipid preparations compared with CAB [I].26;27 If
nephrotoxicity secondary to a lipid preparation did occur, it followed a longer course of therapy
[II].23;25 As rates of drug-induced nephrotoxicity often vary between children and adults, studies
comprising predominantly adults need to be interpreted with caution. Adult and neonatal case
series have demonstrated that lipid preparations are safe in subjects with pre-existing
nephrotoxicity.86-88 In contrast to the findings of Wingard et al,89 a recent meta-analysis
demonstrated no significant difference in rates of nephrotoxicity seen with different lipid
preparations of amphotericin B [I].90 Both azoles and echinocandins cause less nephrotoxicity
than CAB and liposomal amphotericin B in adults and adolescents [II].28;31;33;43;45;60

Risk factors for amphotericin B induced-nephrotoxicity include amphotericin B dose [III-2], pre-
existing renal impairment [III-2], hypovolaemia [II], hyponatraemia [II] and concurrent use of
nephrotoxic medications [II, III-2]. Increasing age [II, III-2] and underlying disease [III-2] are not
associated with an increased risk.23-25;91-97 When independently assessed, cyclosporine and
diuretics increase the rate of nephrotoxicity whereas aminoglycosides and vancomycin in
isolation appear not to increase the risk [III-2].92-94 In predominantly adult subjects with fever and
neutropenia, the risk of nephrotoxicity more than doubled when two or more concurrent
nephrotoxins (cyclosporine, aminoglycosides or foscarnet) were used.24 Avoidance of
nephrotoxins and hypovolemia as well as sodium loading prior to amphotericin use may decrease
the risk of nephrotoxicity. Adult haematopoietic stem cell recipients with aspergillosis are 5 times
more likely to require haemodialysis when administered CAB than solid organ and non-transplant
related chemotherapy recipients.91

A rise in creatinine is frequently seen after starting an amphotericin B preparation.79 Renal
function and electrolytes should be monitored closely in all children administered an amphotericin
B preparation. Mild to moderate nephrotoxicity can be tolerated in many children with serum
creatinine returning to normal upon cessation. The tolerance of amphotericin B-induced
nephrotoxicity is influenced by concurrent co-morbidities and nephrotoxins, the length of therapy
required, degree of nephrotoxicity, ate of deterioration in renal function and the alternative agents
available. These decisions should be made on a case by case basis. (See Box 5,
recommendation 4)

Finalised version: September 2007 10